Kidney International, Vol. 40 (1991), pp.

326—341

NEPHROLOGY FORUM

Dietary protein restriction in patients with chronic renal failure

Principal discussant: WILLIAM E. MITCH

Emo,y University School of Medicine, Atlanta, Georgia

infections. She also had noted swelling of her ankles, especially after
sitting for long periods.
Editors Physical examination revealed a slender, pleasant woman who was
JORDAN J. COHEN well informed about the consequence of renal failure. She weighed 55.5
JOHN T. HARRINGTON kg and her blood pressure was 200/98 mm Hg. Her pulse was 86
beats/mm; respirations, 16/mm; and she had no skin lesions. The
NICOLA0s E. MADIAS physical examination was unremarkable except for I + pedal edema
bilaterally. Serum biochemistry values were: urea nitrogen, 59 mg/dl;
Managing Editor serum creatinine, 2.9 mg/dl; bicarbonate, 27 mEq/liter; potassium, 4.1
CHERYL J. ZUSMAN mEq/liter; calcium, 9 mgfdl; phosphorus, 4.8 mg/dl; uric acid, 7.4 mg/dI;
alkaline phosphatase, 67 U/liter; triglycerides, 131 mg/dl; and choles-
terol, 231 mg/dI. Her hematocrit was 36%, and the white count and
platelet count were normal. Urinalysis showed 1+ protein and 8—10
State University of New York at Stony Brook erythrocytes/high-power field, but no leukocytes or casts.
and The purpose of a restricted diet was discussed with the patient and
her husband, and she decided to begin this therapy. She was evaluated
Tufts University School of Medicine by a dietician who judged her usual intake to include more than 60 g of
protein/day (>1 g/kg/day); she was instructed how to calculate the
protein content of foods. She was given diet plans for a 25 g/day protein
(mixed-quality) and 30 kcal/kg/day diet. Her medications included
Aminess, 5 tablets 3 times daily; calcium carbonate, 3.75 g/day;
Case presentation B-vitamin complex, 1 tablet daily; furosemide, 80 mg/day; and atenolol,
50 mg/day. Over the past 6 years, she has been in regular contact with
A 54-year-old woman from Texas sought a second medical opinion 6 a dietician.
years ago at the Brigham and Women's Hospital in Boston because of During these 6 years, the patient has been treated for hypertension
concern about her family history of renal disease. In 1978, her creati- with several types of drugs, including an angiotensin-converting-en-
nine clearance (Ccr) was reported to be 56 mI/mm and her blood zyme inhibitor, different beta-blockers, and calcium-channel blockers.
pressure 170/100mm Hg. She had been told by her family physician that She has had two bouts of severe diarrhea with dehydration and was
she had hypertension and chronic renal failure and had been given a hospitalized on one occasion for administration of intravenous antibi-
thiazide diuretic. She has an extensive family history of nephritis otics. Radiologic evaluation led to a diagnosis of diverticulosis. Other-
involving at least 4 generations (Fig. I). wise, she has had no symptoms attributable to renal failure. Her current
Between 1978 and 1984, her physician treated her for urinary weight is 56.8 kg. She walks about I mile daily and works as a church
infections and hypertension. In 1981, her Ccr was 42 mI/mm. In 1983, administrator. Serial estimates of her protein intake were calculated
she had a urinary infection with pyuria; her Ccr was 30 mI/mm and a from 24-hour urine specimens for urea nitrogen content using the
hearing test was "normal." In the fall of 1984, she was referred to a relationships described by Maroni et al [1]. Urea nitrogen excretion
nephrologist who noted hypertension and edema. Measurement of averaged 4.8 g N/day, with a coefficient of variation of 17% (N = 44).
glomerular filtration rate (GFR) by 25I-iothalamate renal clearance Renal function was analyzed by measuring serum creatinine, creatinine
disclosed a value of 20.7 mI/mm; propranolol and a thiazide diuretic clearance, and '25l-iothalamate clearance. Values of serum creatinine,
were prescribed. She came to me for another opinion; since then she GFR, and estimated protein intake (uncorrected for the essential amino
has been followed annually, and every 4 to 8 weeks by the referring acid supplements) during the past 6 years are shown in Fig. 2.
nephrologist.
In October 1984, I found the patient to be a well-nourished woman in
no acute distress except for understandable concern about her progno- Discussion
sis for familial renal disease (Fig. I). She had not had a renal biopsy, but
renal biopsy of others in the family had shown an interstitial nephritis; DR. WILLIAM E. MITCH (Director, Renal Division, and E.
none of the family members had any hearing disorder. A review of Garland Herndon Professor of Medicine, Emory University
systems revealed hypertension but no hearing impairment. She had no School of Medicine, Atlanta, Georgia): Early studies revealing
symptoms suggesting cardiopulmonary disease except for mild fatigue that patients with chronic renal failure and severe uremia
on exercising, but she did complain of intermittent constipation. Uri-
nary symptoms included bouts of frequency and urgency occurring improved symptomatically when dietary protein was restricted
about two times a year that had been diagnosed as urinary tract became the initial rationale for dietary therapy [2, 3]. In some
as-yet-unidentified way, the accumulation of nitrogenous waste
products and inorganic ions causes the clinical and metabolic
Presentation of the Forum is made possible by grants from Pfizer, disturbances characteristic of uremia. Dietary protein restric-
Incorporated; Sandoz, Incorporated; Marion Merrell Dow Incorpo- tion has been used as therapy because excess dietary protein (in
rated; Merck Sharp & Dohme International; and Amgen Incorporated. contrast to carbohydrate and fat) is not stored but is degraded to
© 1991 by the International Society of Nephrology urea and other waste products, which accumulate in patients

326
 Nephrology Forum: Protein restriction in CRF 327

11

III

Proband

lv

Deceased with disease

Alive with disease

Deceased )kidney status unknown)

Fig. 1. Clinical observations in 58 members (4 generations) of a family with a hereditary form of renal disease. The disease is inherited in an
autosomal dominant fashion. The patient (proband) has intermittent microscopic hematuria and chronic renal failure.

with impaired renal function. Protein restriction also helps also will address the evidence that dietary manipulation can
control the accumulation of sodium, potassium, phosphates, influence the course of renal insufficiency.
and other ions normally excreted by the kidney because foods
rich in protein contain large amounts of these ions as well. The Urea and protein intake
difficulty in implementing dietary protein restriction therapy is No single toxin has been identified that can account for the
that an inadequate protein intake results in net degradation of spectrum of abnormalities found in uremia. Urea has been the
endogenous protein stores, which contributes to the loss of lean most widely studied because the BUN concentration is related
body mass observed in severely uremic subjects treated with to the severity of uremia. This relationship was underscored
excessively restricted diets [4]. when the National Cooperative Dialysis Study reported that
As the patient discussed demonstrates, it is possible to avoid maintaining the average BUN close to 50 mg/dl was associated
this complication; the treatment she has followed for 6 years with fewer complications than was maintaining the average
yields a protein intake equivalent to about 38 glday. Her weight BUN approximately 90 mg/dl by less-intensive dialysis [5].
has remained stable; she continues to work and exercise Johnson et al tested the toxicity of urea by adding it to the
regularly. She has overcome bouts of diverticulitis on two dialysis bath of otherwise "well-dialyzed" patients [6]. They
occasions. Stabilization of her residual renal function over this found that raising the BUN to about 140 mg/dl for more than
same 6-year period (Fig. 2) also raises the possibility that one week reproduced uremic symptoms. On the other hand, a
dietary protein restriction might slow the rate of loss of renal moderately high BUN causes few symptoms when renal func-
function, even in patients with inherited forms of renal disease. tion is otherwise normal. Patients with defects in their ability to
Prospective trials currently are underway in Europe, Japan, and excrete urea who have moderate azotemia even though their
the United States to test this exciting but unproved possibility. GFR is normal do not have uremic symptoms [7], nor did a
I will discuss the principles underlying the implementation of woman who ate enormous quantities of fishmeal protein and
a low-protein diet in patients with chronic renal failure and will maintained a BUN above 65 mg/dl for at least 3 years [8].
place special emphasis on factors affecting nutritional status. I Regardless of whether urea per se is toxic, the rate of urea
328 Nephrology Forum: Protein restriction in CRF

4—

3— • •. . •. • •• U U
.•.
Serum Cr
mg/dI
2—

8-
6- • •
Urea excretion
g N/day
4- •• •'•
• •• •••.
2-

30 —

20 — £
GFR A£ £ A£
mi/mm L Fig. 2. Serial values of serum creatinine, 24-
10 - L hour urea excretion, and '251-iothalamate
clearance (GFR) in the patient discussed here
0- ___________________________________________________________ with chronic renal failure from hereditary
renal disease. The 24-hour urea excretion
0 10 20 30 40 50 60 70 80
values were used to calculate protein intake
Month as described by Maroni et al [1].

production is directly related to the amount of protein eaten BUN and weight are stable (that is, the patient is in a steady
(Fig. 3). Hence, the BUN concentration is related to protein state), the UNA is equal to urea nitrogen excretion. Calculation
intake, provided urea clearance is stable. This relationship of UNA assumes that short-term variations in body weight
arises because urea is synthesized using nitrogen derived from reflect only changes in body-water content.
amino acid catabolism. The only suggestion that urea metabo- Calculation of UNA can be used in a least three ways to
lism is abnormal in uremic subjects is that some patients have a assess dietary therapy of patients with chronic renal failure.
mildly high arterial ammonia level [9]. The assessments are based on the principle of conservation of
Urea has three fates: it is excreted by the kidney, accumu- mass and the following relationship: BN = 'N — [UNA + 0.031
lates in body water, or is metabolized to ammonia plus carbon gN/kglday]. First, when nitrogen intake (IN) is known, nitrogen
dioxide by gastrointestinal bacteria. In the last case, the ammo- balance (BN) can be estimated accurately because nonurea
nia is recycled to urea in the liver. Consequently, there is no net nitrogen (NUN) excretion does not vary substantially with
loss of nitrogen or urea [10]. The difference between urea protein intake (Table 1). The average value of NUN is 0.031
produced and recycled is termed "urea nitrogen appearance" gN/kg/day (Fig. 4) [1]. Thus, nitrogen intake (16% of protein
(UNA); it is the urea that "appears" in urine and body water. intake) minus the sum of UNA plus the product of 0.031 gN/kg
The UNA is physiologically important because it closely cor- and body weight gives an estimate of BN (Table I). In 19
relates with the amount of nitrogen (and therefore protein) non-dialyzed patients with chronic renal failure consuming 34 to
eaten [1, 11]. Fortunately, UNA can be measured easily. 94 grams of protein/day, BN estimated by this method did not
To measure UNA, the average daily urea nitrogen excretion differ statistically from values obtained by actual measurement
first must be determined. Since urea excretion is not constant of BN [1]. Second, if BN is assumed to be zero, protein intake
throughout the day, shorter collections will not suffice. The can be estimated from UNA using the same relationship (Table
second step is calculating the average change in the urea pool. 1). When the estimated and prescribed intakes differ by more
This calculation is relatively simple because the urea space is than 20%, investigation for occult gastrointestinal bleeding or
equal to body water, assumed to be 60% of body weight in an unrecognized catabolic illness is indicated; or, intensive
nonedematous patients [1, 12]. Because urea distributes equally dietary counseling might be needed. Third, if urea clearance is
throughout body water, the average change in the pool can be known, the steady-state BUN expected for a prescribed protein
calculated from changes in BUN and weight measured over 5 intake can be calculated: if the quotient of UNA in gN/day
days [101. To calculate UNA, the daily rate of change in the divided by urea clearance in liters/day is multiplied by 100, the
urea poo1 (either positive or negative) is added to the average answer is the steady-state BUN in mg/dl (Table 1). Note that in
rate of urea excretion measured over 3 days. Obviously, if the the third example, the steady-state BUN is being calculated
 Nephrology Forum: Protein restriction in CRF 329

Table 1. Relationships among protein intake, nitrogen balance, and

24 steady-state BUN
If the protein intake of a 70 kg patient is 50 g/day and the steady-
state urea excretion (UNA) is 10.3 gN/day, then:
20
BN = 'N - (UNA + NUN)
BN = 12.8 — [10.3 + (0.31 x 70)]
16 BN = 0
y —4.40 - (2.08x)
If nitrogen balance is assumed to be zero, then intake can be
9/day 12 estimated:
BN = 'N - UNA NUN -
8
'N = UNA + (0.31 gN/kglday x weight)
U riflery
urea
If a 70 kg patient has a urea clearance of 8.6 liters/day (6 mllmin),
4
then the steady-state BUN is estimated as:
- 0
0 (IN — 0.031 gN/kg)
0'-
'S 0 steady-state BUN = ,,
( urea)
X 100
I I I I I I
2 Abbreviations: BN, nitrogen balance (gN/day); 'N' nitrogen intake
(gN/day); UNA, urea nitrogen appearance rate (gN/day); BUN, blood
0 urea nitrogen concentration (mg/dl); Curea, urea clearance (liters/day).
—1
Nitrogen balance
9/day
0
0 oQ-4----
Using estimates of UNA and NUN to evaluate nitrogen
+1 metabolism has limitations. First, incomplete urine collection
I , can limit the precision of calculations based on UNA. The only
14 12 10 8 6 4 2 0 way to reduce this error is to collect urine for 3 days and obtain
Nitrogen intake, g/day an average value. Second, urea clearance varies during the day.
Fig. 3. Nitrogen balance and urinary urea as a function of nitrogen This variation could explain why several days are required for
intake in patients with chronic uremia fed varying quantities of dietary normal subjects to reach a new steady-state of urea excretion
protein. Subjects fed less than 4 g nitrogen/day were in neutral or following a change in diet [14]. In patients with renal damage,
negative balance. When the diet contained more than 4 g nitrogen/day, urea clearance may vary less during the day, but more time will
balance was slightly positive, but the increment in intake was equal to
the increment in urinary urea nitrogen. (From Ref. II). be required to achieve a new steady state with stable BUN and
weight. These factors make it imperative to include an estimate
of the change in accumulated urea when assessing UNA in
(thus BUN and weight must be stable), so UNA is equal to urea patients with renal insufficiency; for normal subjects, estimating
excretion. changes in accumulated urea is not critical because of the 7- to
I emphasize these concepts because of their practical utility 8-hour half-life of urea. Even in normal subjects, changes in the
and because I want to clarify some confusing terminology. size of the pool do not fully explain the delay in achieving a
Similar concepts were used by Sargent et al to describe urea steady state following changes in dietary protein. Perhaps
generation (Ga) and the protein catabolic rate (PCR) [13]. In differences in protein utilization (and hence, waste nitrogen
dialysis patients as well as nondialysis patients, the nitrogen for production) occur in the presence of large day-to-day changes
urea synthesis is derived from amino acid degradation. In short, in dietary protein.
net urea production (or G) in dialysis patients is the same as Third, proteinuria can affect accurate measurement of nitro-
UNA. Because G and UNA are calculated from the same gen metabolism. The 0.031 gN/kg/day value for non-urea nitro-
variables, they both closely parallel nitrogen (and therefore gen excretion was derived from patients with urinary protein
protein) intake. The more confusing term, PCR, is simply the excretion less than 5 g/day (Fig. 4). If proteinuria exceeds 5
protein equivalent of the UNA plus a constant for non-urea glday, the extra nitrogen lost must be added to the NUN value.
nitrogen excretion of 1.7 gN/day. As I said, PCR closely Evidence indicates that proteinuria affects protein balance.
approximates protein intake in patients who are in nearly zero Kaysen and associates showed that in nephrotic rats, protein-
nitrogen balance (BN). Clearly, PCR does not measure protein uria decreased lean body mass, and raising dietary protein
exacerbated proteinuria [15]. In humans, the same group of
catabolism because the daily turnover of protein in humans is
investigators showed that reducing dietary protein to 0.8 g/kg
far greater than protein intake (and hence, PCR). Results of body wt/day from 1.6 g/kg/day resulted in less proteinuria
isotope dilution studies place both the rate of protein synthesis without compromising albumin homeostasis [16]. Even though
and of degradation at 45—55 gN/day [14], roughly equivalent to a protein-restricted diet can spare albumin stores, its effect on
the protein in 1.0 to 1.5 kg muscle. Although the principle of body protein stores is less clear, especially during long-term
conservation of mass dictates that the difference between therapy.
"whole-body" protein synthesis and degradation must equal The protein requirement, defined as the minimum amount of
waste nitrogen production x 6.25, the implication that PCR protein that balances obligatory nitrogen losses while maintain-
yields insight into, or is a measure of, whole-body protein ing energy equilibrium and permitting modest physical activity,
catabolism is erroneous. can be divided into two components: total nitrogen and essen-
330 Nephrology Forum: Protein restriction in CRF

50

40 lj
30 l!l

z E
20
*

Fig. 4. Calculated values of total non-urea

10
nitrogen excretion (NUN) in normal subjects
(A, •, U), and patients with chronic renal
failure being treated with nutritional therapy
0 (•, ®. ®,®) by dialysis (•, ti), or CAPD
2 4 6 8 10 12 14 16 18 20
l). The average value found in the study
Nitrogen intake by Maroni et at was 0.031 g nitrogen/kg/day
gN/day (©). (From Ref. 1.)

tial amino acids (EAA). From these considerations, a pre- a shorter half-life than albumin (about 8 days versus 20 days for
scribed diet may be deficient in protein quantity, quality (that is, albumin). However, transferrin concentration, like albumin,
the amount of EAA), or both. Despite its limitations, nitrogen varies with hydration and protein turnover but also changes in
balance (BN) remains the "gold standard" for assessing dietary the presence of inflammation and/or changes in iron stores [18,
protein requirements [14]. To measure BN, the difference in 20].
nitrogen in the diet and that in urine and stool plus an "allow- Besides measurements of BN and serum proteins, anthropo-
ance" (for example, 0.5 gN/day to account for minor losses metrics (weight, mid-arm muscle circumference, triceps skin-
from sweat, desquamated skin, or hair) is obtained. For uremic fold thickness) are commonly used to estimate body composi-
patients, the urea pool also must be taken into account. tion and nutritional adequacy [18]. Reproducibility depends on
When defining nitrogen requirements, one must specify an the skill of the observer, the number of sites examined, the use
adequate energy intake, because for any given protein intake, of the dominant or non-dominant arm, and the degree of
BN improves when calories are increased [17]. It is important hydration (for example, in dialysis patients). Unfortunately,
that we understand how current protein recommendations were many of the values used as standards were obtained from
established: they are based on interpolation of several measure- studies in underdeveloped countries, where the health of the
ments of BN in individuals to identify the level of dietary protein study population was not defined and may not represent a
that should produce zero BN for groups of subjects. Because standard for normal values. Also, the influences of age and
the recommended intake for all subjects is based on extrapola- socioeconomic status have not been evaluated. Finally, virtu-
tion of data from individuals, the recommended intake doesn't ally no data have examined how closely subnormal anthropo-
always apply to groups of patients differing in age, degrees of metric values relate to an adverse clinical outcome. Thus, serial
activity, etc.; clearly the protein required by an individual could anthropometric measurements in the same individual are more
vary substantially from the minimum value. For these reasons likely to provide useful information.
the Food and Agriculture Organization/World Health Organi- The concentrations of plasma amino acids also have been
zation/United Nations University (FAO/WHO/UNU) "expert used to assess nutritional status [20]. But the amino acid pattern
group" introduced the concept of a safe intake level [17]. This generally reflects the disease process, recent changes in protein
safe level is defined as the average requirement plus two intake, or both, and is a less sensitive indicator of body
standard deviations, a value that should meet the protein composition than are serum proteins and anthropometrics. On
requirements of all but 2.5% of individuals. the other hand, measurements of plasma amino acids have
Serum albumin and transferrin levels are used extensively to revealed abnormalities that seem to be characteristic of uremia.
assess nutritional status [18]. Experience with other serum For example, the tyrosine level is low because uremia limits the
proteins is too limited to recommend their use. When using activity of the enzyme converting phenylalanine to tyrosine
serum albumin, one should recognize that its concentration is [20]. Also, metabolic acidosis activates skeletal muscle
determined by many factors, including the synthesis, catabo- branched-chain amino acid dehydrogenase, the rate-limiting
lism, and excretion of albumin; changes in plasma volume; and enzyme regulating breakdown of these amino acids [21, 221. In
distribution between extracellular and intravascular spaces. In fact, the subnormal levels of branched-chain amino acids
malnourished nonuremic subjects, as well as in chronically present in rats with chronic renal failure could be attributed to
uremic and nephrotic patients, both albumin synthesis and metabolic acidosis [21]. Recent data from Bergstrom and asso-
catabolism can be decreased; in addition, evidence suggests ciates suggest a similar adverse effect of acidosis in humans;
that albumin shifts into the intravascular compartment [16, 18, these authors documented a close correlation between serum
19]. Clearly, the plasma albumin concentration need not accu- bicarbonate and the concentration of valine in muscle of dialy-
rately reflect total albumin mass under these circumstances. sis patients [231. Plasma amino acid levels might provide insight
Serum transferrin is another indicator of malnutrition; it has into methods of improving nutrition. Alvestrand et al studied
 Nephrology Forum: Protein restriction in CRF 331

Uremia

+ Protein synthesis Abnormal energy metabolism Acidosis

4 Protein degradation 4 Amino acid oxidation

Fig. 5. Factors causing loss of lean body mass and
increased essential amino acid (EAA) and nitrogen
requirements in uremia. Results from experimental
animals with renal insufficiency indicate that both
abnormal energy metabolism and metabolic acidosis
compromise nitrogen metabolism; there is also
Loss of lean body mass evidence that an unidentified factor in uremia limits
Increased EAA and nitrogen requirements the capacity for protein synthesis.

predialysis patients fed 16 to 20 g of protein plus a supplement technique requires a constant infusion of a labeled amino acid
of EAA in proportions designed to correct abnormalities in and collection of expired carbon dioxide to allow calculation of
amino acid concentrations [24]. The regimen substantially cor- rates of protein synthesis and breakdown as well as amino acid
rected intracellular amino acid concentrations and improved catabolism [29]. Normal subjects primarily respond to a sharp
BN. change in dietary protein by varying amino acid oxidation.
When protein intake is below the minimum requirement, how-
Protein conservation in chronic uremia
ever, additional responses are activated, including a decrease
In rats with chronic renal failure, abnormalities in protein both in protein synthesis and degradation [30, 31]. Slowing of
metabolism include increased protein and amino acid degrada- protein turnover blunts the adverse effects of an inadequate diet
tion and impaired insulin-stimulated protein synthesis in muscle on protein stores but it does not prevent loss of body protein.
(Fig. 5) [25]. The catabolic effects of uremia are accentuated by The importance of understanding protein nutrition is that aci-
fasting [26]. To date, the only factor in uremia shown to impair dosis in uremia could increase amino acid oxidation and protein
protein metabolism is metabolic acidosis. Even mild metabolic breakdown and thus limit a patient's capacity to successfully
acidosis (serum bicarbonate less than 21 mM) can account for adapt to a low-protein diet.
accelerated catabolism of protein and branched-chain amino Information about how patients with chronic renal failure
acids in muscle of rats with chronic renal failure. Both catabolic adapt to a low-protein diet is limited. Goodship and colleagues
processes were fully corrected when sodium bicarbonate was used the amino acid infusion technique in patients whose
added to the diet [20, 25]. The same adverse effects of acidosis
average serum creatinine was 5 mg/dl and in normal subjects.
are present in humans with chronic renal failure; sodium
bicarbonate decreases urea production and improves BN in Both groups were fed two levels of protein, 1.0 and 0.6 g
these patients [27]. As I noted, evidence indicates that acidosis protein/kg/day [29]. They tested the subjects' capacity to adapt
changes branched-chain amino acid metabolism in muscle of to the minimum protein intake (0.6 g/kg/day) by measuring
patients with chronic renal failure [23]. Together these results whole-body amino acid oxidation and protein synthesis and
suggest that the metabolic acidosis of chronic renal failure degradation; they also measured BN. The authors reached two
increases nitrogen requirements and could counteract adaptive important conclusions: (1) The metabolic responses to the
responses to a low-protein diet. low-protein diet of patients with chronic renal failure and
Energy metabolism also influences protein metabolism in normal subjects were indistinguishable. Both groups had re-
chronic renal failure (Fig. 5). In rats with chronic renal failure, duced amino acid oxidation and protein degradation; protein
adiposity is subnormal and inversely correlates with body synthesis changed minimally with the low-protein diet. How-
weight and muscle protein turnover [26]. Besides the evidence ever, none of the patients was acidotic, so more severely
that insulin-mediated glucose metabolism is abnormal in ure- uremic (or acidotic) patients might have different responses
mia, muscle lactate production correlates highly with protein from normal subjects. For example, acidosis could stimulate
turnover [27]. protein breakdown and reduce the efficiency of using dietary
The metabolic responses that permit adaptation to a protein- protein by stimulating essential amino acid catabolism. (2) The
restricted diet were identified only recently. The problems of normal subjects and the patients were in negative BN during the
assessing "whole-body" protein synthesis and degradation and first week of the low-protein diet (Fig. 6). This response could
interpreting how changes in these values are expressed in not be explained by a low energy intake or by changes in the
specific tissues continue to perplex investigators. Moreover, urea pool (the latter was measured). It is interesting that the
there are large voids in our understanding of factors controlling nitrogen balance occurring with a low-protein diet was dis-
protein metabolism [281. tinctly different from that of the patients fed even less protein
Changes in whole-body protein metabolism have been exam- but given a supplement of ketoacids [32]. In patients fed this
ined by analysis of the turnover of infused amino acids. The regimen, BN was zero.
332 Nephrology Forum: Protein restriction in CRF

3
Fig. 6. Nitrogen balances measured using
classical and isotopic methods in patients
2 with chronic renal failure (CR1) and in normal
subjects fed I g protein/kg/day (HP) and 0.6 g
protein/kg/day (LP). Isotopic nitrogen balance
gN/day was calculated from rates of protein synthesis
0 and degradation measured during infusion of
L-[ l-C] leucine and extrapolated to
represent rates over 24 hours. The difference
is the nitrogen balance during the HP period
minus that in the LP period for CR1 patients
—2 and normal subjects. The difference in
measured or calculated nitrogen balances was
the same for both groups of subjects. (From
HP LP HP LP Measured "Isotopic" Ref. 29.)

Dietary protein requirements in uremia promoting neutral BN and muscular strength and maintaining
More than 25 years ago, studies suggested that patients with serum albumin and transferrin [24]. In another study, serum
advanced chronic renal failure could maintain BN while con- albumin and transferrin were maintained and lean body mass
suming very low amounts of protein (about 4 SD below the increased or remained stable as long as no intercurrent illnesses
mean requirement for normal subjects) [2, 3]. Giordano postu- prevented an adequate caloric intake [36].
lated that nitrogen derived from urea degradation could be used Essential amino acids usually are provided in the proportions
to synthesize amino acids and ultimately protein [2]. Tests of recommended for normal subjects. It is likely, however, that
this hypothesis in patients with chronic renal failure have failed these proportions are not optimal for patients with chronic renal
to confirm that urea reutilization contributes significantly to failure, because plasma and intracellular concentrations of
protein nutrition. For example, only 2% to 6% of the nitrogen in amino acids in uremic patients differ markedly from those in
albumin from patients with chronic renal failure was derived normal subjects or normal subjects eating a low-protein diet
from urea degradation [33]. Furthermore, when we suppressed [20].
urea degradation by administering nonabsorbable aminoglyco- A newer EAA supplement (Aminess-Novum, KabiVitrum)
side antibiotics orally, analysis of urea kinetics showed that contains tyrosine. Tyrosine's synthesis from phenylalanine is
ammonia derived from urea degradation was simply recycled to reduced in uremic patients, and giving them tyrosine seems to
urea [10] and BN actually improved [34]. Clearly, if ammonia improve their nutrition [20]. Histidine is included in this new
arising from urea degradation is an important source of nitrogen supplement because these patients given a histidine-deficient
for amino acid synthesis, BN should have become negative diet developed a syndrome characterized by malaise, an erythe-
when urea degradation was inhibited. We concluded that urea matous scaling rash, and negative BN; supplemental histidine
degradation does not contribute substantially to nutrition in rapidly corrected these abnormalities [20]. The new EAA
patients with chronic renal failure. supplement also contains a higher proportion of valine. During
The FAO/WHO/UNU expert group concluded that the aver- long-term therapy, intracellular amino acid concentrations im-
age protein requirement for normal subjects is about 0.6 g proved, as did BN [24]. These results strongly suggest that the
protein/kg/day [17]. It is now recognized that predialysis pa- amino acid requirements of uremic subjects are different from
tients require about the same amount as normal subjects. To those of normal individuals.
ensure an adequate intake of EAA, approximately two-thirds of Nitrogen intake also can be reduced by taking advantage of
the protein should be of "high biologic value" (for example, the capacity of most tissues to convert the alpha keto analogues
eggs or lean meat) and calories must be sufficient to meet energy of seven EAA (ketoacids) to EAA. Ketoacids of the two
requirements [35]. These recommendations are emphasized remaining EAA, lysine and threonine, are not transaminated,
because of reports that this level of protein is associated with and these amino acids must be provided. Proof that a mixture of
negative BN, at least during the first week of adaptation to the ketoacids could replace their respective amino acids in the diet
diet [291. of patients with chronic renal failure was provided by showing
More severe protein limitation than the minimum require- that BN was zero over periods of 5 to 7 days even when patients
ment of 0.6 g protein/kg/day is ill advised because EAA intake were eating a virtually protein-free (and hence, EAA-free) diet
would be insufficient to maintain BN. An acceptable alternative [37].
is further limitation of protein (about 20 glday) and the addition In early studies, ketoacids were provided to uremic patients
of a supplement of EAA or their nitrogen-free analogues as calcium salts [20]. Gastrointestinal distress was an occa-
(ketoacids). This strategy actually could increase the variety ofsional problem, and about 5% of patients developed hypercal-
foods and improve compliance because all protein does not cemia during long-term therapy, although the latter was rarely
have to be of the "high-quality" type. Generally, a regimen of clinically significant. A new supplement was designed to avoid
15 to 25 g/day of unrestricted quality protein plus EAA provided the problems associated with calcium salts while providing a
in capsules or tablets promptly corrects uremic symptoms while more optimal proportion of ketoacids and amino acids for
 Nephrology Forum: Protein restriction in CRF 333

uremic patients [32]. Ketoacids of the branched-chain amino acids. Lessening the catabolic influence of acidosis might
acids are given as salts of basic amino acids (ornithine, lysine, partially explain the beneficial effect of restricted diets on
and histidine). Tyrosine, small amounts of threonine, and the protein metabolism in renal failure [21, 25, 27]. In summary,
hydroxy analogue of methionine are added, but phenylalanine long-term treatment of patients with severe renal insufficiency
and tryptophan are omitted because the diet contains enough of with a very-low-protein diet supplemented with EAA or ketoac-
these amino acids to achieve nitrogen balance [32]. ids maintains nutrition while reducing uremic symptoms, sec-
A diet containing 20 to 25 g of mixed-quality protein plus this ondary hyperparathyroidism, and metabolic acidosis.
mixture of ketoacids produced a zero BN and decreased UNA High-protein diets have been used for nephrotic patients but,
in patients with severe chronic renal failure (average GFR, 4.8 in view of recent reports, the appropriateness of this maneuver
ml/min) [32]. Weight and serum albumin and transferrin levels is questionable [161. When albumin turnover and excretion
were maintained over periods ranging from 4 to 19 months [32]. were measured in 9 nephrotic subjects fed a high- (1.6 g/kg/day)
These results are particularly interesting because the supple- and low- (0.8 g/kg/day) protein diet containing 35 kcal/kg/day
ment was "unbalanced"; that is, it lacked certain EAA or EAA for 10 to 14 days, the more restrictive diet was associated with
analogues. It is important that neither the ketoacid nor an EAA a significant reduction in albumin excretion (mean, —2.74 g/day;
supplement has any detectable nutritional benefit in uremic range, —0.1 to —7.4 g/day) and with a small increase in serum
subjects eating more than 40 g/day of predominantly high- albumin (+0.2 g/dl). Although albumin synthesis was lower
quality protein, presumably because the excess EAA or ana- with the 0.8 g/kg/day diet, this decrease was more than com-
logues are metabolized [201. pensated for by decreases in albumin catabolism and fractional
Lucas et al suggested that a ketoacid regimen can reduce lean excretion. Albumin synthesis did increase significantly on the
body mass [191. They treated 12 patients with only 0.2 g high-protein diet but was offset by increased proteinuria. In
protein/kg/day (versus approximately 0.3 g/kg in other studies) spite of these encouraging results, simultaneous measurements
and with approximately 50% of the amount of ketoacids used by of BN and other nutritional indices are needed to assess body
others [20, 32]. Therapy was associated with a decrease in protein stores before protein-restricted diets can be recom-
symptoms as well as a decrease in BUN, serum phosphate, mended for all nephrotic patients. Moreover, protein-depleted
parathyroid hormone, and the calcium-phosphorus product. subjects and children might require more dietary protein.
Despite stable serum albumin and transferrin levels, the body
weight, creatinine excretion, and mid-arm muscle circumfer- Energy requirements in uremia
ence decreased significantly, suggesting loss of lean body mass. The energy requirement is defined as the calories needed to
This study emphasizes the importance of providing a sufficient maintain health and normal physical activity. Because inade-
intake of protein and ketoacids (and possibly energy); it also quate calories, defective energy metabolism, or both could raise
points out the hazards of monitoring changes in plasma proteins protein requirements and aggravate uremia, successful therapy
as the sole index of nutritional status. requires that caloric intake be adequate. Unfortunately, little is
A possible anabolic effect of ketoacids is supported by known about factors influencing energy requirements or the
studies suggesting that alpha ketoisocaproate (the ketoacid of complex relationship between caloric intake and maintenance
leucine) improves protein conservation. In-vitro studies using of protein stores.
rat muscle show that alpha ketoisocaproate reduced protein The FAO/WHO/UNU recommendations for energy intakes
degradation [38]. Evidence also points to an anticatabolic effect were determined from approximately 11,000 basal metabolic
in humans. Infusion of alpha ketoisocaproate into starving, rates in healthy individuals of both sexes and all ages. Although
obese adults with normal renal function decreased urea nitrogen these recommendations are considered the best available, it
production, whereas an equimolar infusion of leucine did not should be recognized that as much as 50% of the variability in
change BN [39]. The clinical relevance of alpha ketoisocaproate energy requirements must be attributed to factors other than
to dietary therapy is presently unclear, and further investigation age, gender, and weight [17]. With regard to designing a diet for
is warranted. a uremic patient, the precision of estimating the energy require-
A ketoacid regimen also might improve clinical and biochem- ment depends heavily on the consistency and accuracy of
ical evidence of secondary hyperparathyroidism. For example, estimated time spent on various physical activities.
ketoacid therapy treatment has been associated with decreased In patients with chronic renal failure, adaptation to inade-
serum phosphate, alkaline phosphatase, and parathyroid hor- quate caloric intake could involve changes in metabolism,
mone, as well as with increased serum calcium and improve- decreased physical activity, and/or loss of lean body mass (Fig.
ment in renal osteodystrophy [20]. Serum levels of I ,25-dihy- 5). In semi-starved subjects, the basal metabolic rate decreased
droxy-cholecalciferot increased, whereas serum concentrations about 15% over 3 weeks; further adaptation reduced lean body
of 25-hydroxy-cholecalciferol and calcitonin did not change mass [17]. Well-nourished individuals can achieve energy bal-
[20]. Although the improvement in divalent ion metabolism is ance with only one-half of their usual calories if physical
likely due in part to the concomitant reduction in dietary activity is limited, but lean body mass will diminish. Evidence
phosphorus associated with this regimen, suppression of phos- suggests that the range of caloric intake compatible with suc-
phate absorption also might contribute; calcium salts of ketoac- cessful adaptation is small [17].
ids appear to be at least as effective as calcium carbonate in Few studies have examined the caloric requirements of
binding phosphorus [20]. patients with chronic renal failure or how they adapt to a
Improvement in metabolic acidosis is an obvious conse- reduced caloric intake. Kopple et at varied the energy intake of
quence of the reduction in hydrogen ion produced during 6 pre-dialysis patients between 15—45 kcallkg/day while keeping
metabolism of dietary phosphate and sulfur-containing amino protein intake at 0.55—0.60 g protein/kg/day. The authors con-
334 Nephro!ogy Forum: Protein restriction in CRF
cluded that 35 kcal/kg/day could adequately maintain BN [351.
The same group found that energy expenditure of predialysis
and hemodialyzed patients during rest and exercise did not decline in the reciprocal of serum creatinine (Scr) over
differ from that of control subjects [401. Thus, calorie require-
ments of stable patients with chronic renal failure, like protein
requirements, seem to be similar to those of normal subjects
[20].
Progression of chronic renal failure
Patients with established renal failure rarely recover; they
continue to lose renal function even when the disease process
that initially damaged the kidneys is no longer active [20]. In the
early 1970s, it was believed that the natural history of renal
insufficiency was unpredictable; the outcome of renal disease
usually was described as the percentage of patients surviving
until end-stage renal disease occurred or until a predetermined
level of serum creatinine was reached. In fact, the rate of loss of
renal function over time is constant in the majority of patients,
although its magnitude can vary widely among individuals with
the same disease. To what degree the severity of the underlying
disease process, other concomitant factors such as hyperten-
sion, or the method used to assess renal function contribute to
variability in rates of progression is unknown.
Accurate assessment of the natural history of progressive
renal insufficiency is useful both for prognosis and for deter-
mining whether therapy has altered the course of the underlying
disease. To evaluate the efficacy of a specific therapy and/or
factor(s) influencing the loss of residual renal function, two
questions must be asked: (1) How can residual renal function be
estimated; and (2) How can changes in renal function be
measured most precisely? These questions address different
tasks; the former determines renal function at one point in time,
whereas the latter examines how renal function changes during
the course of the illness.
The glomerular filtration rate has been estimated from serum
creatinine. However, a single value of serum creatinine is only
a crude estimate of the glomerular filtration rate; serum creat-
mine might not be elevated above the 95% confidence interval
of "normal" until the GFR has declined by at least 50% [411. As
Levey discussed in a recent Nephrology Forum devoted solely
to the issue of measurement of GFR, variability in creatinine
production and tubular secretion results in wide confidence
intervals for the relationship between serum creatinine and
inulin clearance [42]. As renal failure advances, creatinine
excretion decreases due to extrarenal creatinine clearance [43].
Extrarenal clearance was found to be relatively constant,
averaging 0.038 liter/kg/day; this finding indicates that as renal
function declines, extrarenal metabolism represents a progres-
sively greater proportion of total creatinine elimination. Several
factors influence creatinine production, so it is not surprising
that questions have arisen about its use as a measure of renal
function. There also is the problem of day-to-day variability in
24-hour creatinine clearances [44, 45]. These concerns limit the
use of serum creatinine and creatinine clearance as a measure-
ment of GFR and, by implication, the number of functioning
nephrons. Although this topic has been reviewed, the published
data comparing methods are scarce [41, 42, 44, 45].
One major problem in analyzing whether changes in diet alter
the progression of renal disease is determining the predictability
of repetitive measurements of renal function. The realization
that the decline is predictable and that it occurs at a constant
rate was first recognized in 1976, when it was reported that the
periods averaging 71 months was linear in 31 of 34 patients [46].
Patients in this initial report had not had any dietary manipula-
tion. At least two lines of evidence suggest that changes in the
clearance of creatinine and Scr' might not reflect changes in
GFR as accurately as the renal clearance of radiolabeled
glomerular markers. First, fractional tubular creatinine secre-
tion was found to vary inversely with GFR in patients followed
longitudinally [41]. In patients whose GFR improved by about
33%, Shemesh et al found that changes in creatinine clearance
or Scr suggested only about a 13% improvement. Of 26
patients who underwent a remission (defined as an increase in
GFR and a decrease in albuminuria), serum creatinine did not
decline in 14, and only 13 of the 26 had an increase in creatinine
clearance. On the other hand, serum creatinine increased in all
but 3 of 28 patients with declining function, whereas creatinine
clearance declined in 22 of 28 patients. Thus, changes in
creatinine clearance and Scr' seemed to be more sensitive in
detecting a decline in GFR than in identifying an increase in
GFR. These results emphasize that variability between GFR
and serum creatinine is likely to be greatest in individuals with
modest reductions in GFR.
Second, Walser and coworkers compared changes in creati-
nine clearance and SCr' (corrected for an average value of
urinary creatinine) with the renal clearance of mtechnetium
diethylene triaminopentaacetic acid (99mTc..DTpA) in 17 pa-
tients with moderate to severe chronic renal failure (GFR, 4 to
23 mI/mm) [451. Over an average of 15 months, the slope of
creatinine clearance with time declined more rapidly and had
greater variance than did the renal clearance of 99mTcDTPA.
Although the authors concluded that 9 of 22 observations
revealed differing slopes between changes in creatinine and
99mTcDTPA clearances during all or part of the observation
period, the choice of periods analyzed is sensitive to observer
bias. Over the entire observation period, 4 of 22 observation
periods revealed different slopes for the two methods, and 3 of
those 4 periods suggested progression by creatinine clearance
when none was seen by 99mTc..DTpA In the remaining case,
the isotope clearance suggested improvement while creatinine
clearance remained stable. Interestingly, in 2 of 4 patients
whose slopes revealed stable 99mTcDTpA clearances but pro-
gression by creatinine clearance, dialysis was required; thus, in
those 2 patients, creatinine clearance was a better measure of
clinical outcome than was mTcDTPA clearance.
When rates of progression, as defined by changes in l/Scr,
were compared with mTcDTPA clearance, the variances
were similar, and the average slope for the groups did not differ;
6 of 22 slopes were statistically different between the two
methods when the entire observation period was analyzed. In
two instances, mTcDTPA renal clearance remained stable,
whereas SCr' declined. In two cases, both methods showed
progression (but the rate was statistically greater with the
method). In two instances each, Scr remained stable but
99rnTc..DTPA improved or decreased respectively. Thus, in only
4 of 22 patients would an inappropriate conclusion regarding
directional changes in renal function have been reached by
utilizing Scr. These data support the original proposal that
renal function can be monitored and is lost at a constant rate.
 Nephrology Forum: Protein restriction in CRF 335

Regardless, decisions about the effect of the diet on progressionpolycystic kidney disease. Initial serum creatinine, proteinuria
require more than changes in serum creatinine concentration. on presentation, and systolic and diastolic blood pressures were
Fortunately, the Modification of Diet in Renal Disease (MDRD) found to be independent prognostic factors. No adverse effects
trial uses '251-iothalamate to measure GFR, and this practice of dietary therapy were noted, and indices of protein nutrition
will avoid the problem in that study [47J. were well maintained [48]. However, after an additional 5 years
After considering the variability in rates of progression of dietary restriction, the researchers noted significant loss of
among individuals and the standard error of the slopes of their muscle protein and a decrease in serum albumin and serum
regression lines, Walser concluded that at least four measure- transferrin concentrations (despite stable body weight and
ments of GFR would be necessary to detect progression in most indices of muscle mass) in a subgroup of 8 patients; this finding
patients and that little would be gained by requiring more than suggests that ml' ;tional status tends to worsen after 5 or more
five measurements [44]. Only the most rapid rates are detect- years [52]. Unfortunately, the energy intake of these 8 patients
able with just three measurements. Consequently, estimates of was lower than that prescribed (26—29 kcal/kglday), so it is not
progression based on two measurements (an initial and final clear that dietary protein restriction alone causes protein wast-
value) cannot establish progression rates in individuals unless ing.
the fall in GFR greatly exceeds the error in GFR measurements. In 1984, Rosman and coworkers reported the results of their
Interestingly, as GFR declines, so does its error; hence, pro- prospective randomized trial involving 149 patients followed for
gression becomes easier to detect [20]. at least 18 months (average of 24 months) after assignment to
During the last 15 years, evidence has suggested that dietary low-protein or control diets [49]. The prescription depended on
therapy can slow the progression of chronic renal failure. Many the degree of renal insufficiency: 0.6 or 0.4 g/kg/day of protein
of these studies have been criticized because of study design for patients with creatinine clearances between 30—60 mI/mm or
(including the lack of randomization), retrospective analysis, 10—30 mI/mm, respectively. They concluded that a low-protein
compliance issues, and/or the use of changes in the serum diet significantly slowed the increase in serum creatinine, and
creatinine or creatinine clearance to assess progression. Con- that patients under 40 years of age progressed more rapidly than
sequently, conclusions regarding the efficacy of dietary protein did older subjects. The authors noted no adverse influence of
and/or phosphorus restriction on progression of chronic renal protein restriction on nutritional status. These authors subse-
failure in humans must be considered tentative. quently reported a 4-year followup of 153 of 248 patients
Nonetheless, three nutritional regimens have been prescribed entering the study [53]. Although a significant benefit of dietary
to slow progression: (I) a conventional low-protein diet con- protein restriction still was noted, it was most apparent in the
taining 0.6 g protein/kg body weight/day of primarily high- group with more advanced disease. In both the control and
quality protein; (2) a very-low-protein diet supplemented with low-protein diet groups, males showed a more rapid loss of
essential amino acids; and (3) a very-low-protein diet supple- creatinine clearance but also seemed to respond to protein
mented with ketoacids. restriction; females did not seem to benefit from dietary modi-
The principal evidence that the use of unsupplemented, fication. Furthermore, slowing of disease progression was evi-
low-protein diets can favorably influence the course of chronic dent only in patients with glomerulonephritis. Disease progres-
renal failure is derived from three reports [48—50], two of which sion in patients with polycystic kidney disease appeared to be
were randomized, prospective trials. In the earliest study, entirely related to blood pressure, whereas in the other diag-
Maschio et al in Verona compared the rates of progression in nostic groups, variability in blood pressure did not correlate
three groups of patients [48]. Groups land!! were separated on with preservation of renal function. Finally, body weight and
the basis of initial serum creatinine value, and each received a serum proteins were stable over 36 months of dietary therapy,
diet containing 0.6 g/kg of predominantly high-quality protein, but prescribing less than the minimum daily requirement of
40 kcal/kg energy intake, about 650 mg of phosphorus, and 1.0 protein to patients with a creatinine clearance less than 30
to 1.5 g of calcium daily. The initial serum creatinine in Groups mI/mm could be hazardous nonetheless [17].
1(25 patients), 11(20 patients), and III (30 patients) were 1.5 to In Australia, IhIe et al compared a diet containing 0.4 g
2.7 mgldl, 2.9 to 5.4 mg/dl, and 1.6 to 4.7 mg/dl, respectively. protein/kg/day with an unrestricted protein intake in 64 subjects
Group III was a control population consuming an unrestricted who were followed for 18 months in a prospective, randomized
diet with average daily intakes of protein, phosphorus, and trial [50], The groups were initially well matched for blood
calcium of 70 g, 900 mg, and 800 mg, respectively. Progression pressure, serum creatinine (range, 4.0 to 11.0 mgldl), and
was assessed by changes in serum creatinine or Scr. Groups calcium and phosphorus values. Changes in GFR were deter-
I and II had a far slower loss of renal function than did the mined from the plasma disappearance of 51Cr-EDTA. End-
control group; the rate of decline was significantly slower in the stage renal failure developed in 9 of 33 patients (27%) who
patients who began treatment at an earlier stage of disease. followed the unrestricted diet compared with only 2 of 31(6%)
The Verona group has periodically updated their experience. who were compliant with the protein-restricted diet (P < 0.05);
In 1989 they reported on 390 patients treated with a low-protein the GFR decreased on average from about 15 mI/mm to 6
diet for 54 28 months: 57% of the patients had a stable serum mI/mm in the former group, whereas it did not change signifi-
creatinine level, 11% had slower renal deterioration (defined as cantly in the protein-restricted group (from approximately 14
an increase in Scr between 0.02 and 0.04 mgldllmonth and 32% mI/mm to 12 mI/mm). The outcome of patients who did not
had rapid renal deterioration (greater than 0.04 mg/dl/month) comply with the restricted diet was not detailed. Although
[51]. Individuals with milder renal disease seemed to have a serum albumin and estimates of muscle mass remained stable
more favorable course, and patients with interstitial nephritis over the 18-month followup, weight, serum transferrin, and
fared better than did those with chronic glomerulonephritis or total lymphocyte count decreased significantly. Because the
336 Nephrology Forum: Protein restriction in CRF

phosphorus content of the protein-restricted diet was approxi- unsupplemented diet. In patients compliant with the ketoacid
mately 30% to 40% less than that of the unrestricted diet, the regimen, the decline in creatinine clearance was halted (0.1
relative importance of dietary protein versus phosphorus re- 0.12 mI/mm). Patients who were less compliant (average UNA,
striction on progression cannot be determined. The decline in 6.3 gN/day) continued to lose renal function. We evaluated a
some, but not all, nutritional indices also raises concern about regimen containing 20 to 30 g of protein supplemented with a
prescribing less than the minimum daily requirement of protein mixture of the basic amino acid salts of ketoacids in 25 patients
to patients with renal insufficiency. Clearly, we also need more [57]. Among 17 patients who demonstrated well-defined rates of
sensitive methods for assessing the nutritional adequacy of disease progression, 10 (59%) exhibited a significantly slower
currently recommended dietary regimens. rise in the serum creatinine level during long-term treatment
How did very-low-protein diets supplemented with essential (average, 20 months). Seven of these 17 patients began treat-
amino acids affect the progression of chronic renal failure? ment before the serum creatinine reached 8 mg/dl; in 6, the
Alvestrand and associates treated 17 patients who had well- serum creatinine remained at or below the level at the start of
defined rates of decline of Scr' before beginning a regimen of treatment. Thus, it appears that this regimen has a more
15—20 g of mixed-quality protein plus an EAA supplement favorable influence if initiated relatively early in the course of
provided as tablets (containing 1.8 to 2.8 g nitrogen). The renal failure.
patients were followed for an average of 355 days. Progression All these studies suggesting a beneficial effect of protein
was apparent even though many patients had been consuming a and/or phosphorus restriction are based on creatinine measure-
diet restricted to 0.6 g protein/kg/day [24, 54]. Only 3 patients ments, which can be unreliable estimates of progression. These
did not show substantial slowing of progression. Interim reports results also have not adequately compared a ketoacid-based
from an ongoing prospective, randomized trial by the same regimen with an EAA regimen or with a low-protein, unsupple-
group have cast doubt on what influence diet exerts on these mented diet. Walser and colleagues evaluated 12 patients given
results and raises the possibility that better blood pressure a regimen containing 0.3 g protein/kg/day plus EAA; all had
control and closer followup slowed progression of renal disease progressive decline in creatinine clearance [58]. After the EAA
[55]. Slowing of progression appeared to be related to a very were changed to a ketoacid supplement, renal failure in all 6
small, but significant, reduction in diastolic blood pressure (2 patients whose serum creatinine level exceeded 7.5 mg/dl
mm Hg); loss of function was greater in patients with protein- continued to progress. In contrast, 6 of 7 patients whose serum
uria. The relationship with blood pressure appears to exist at creatinine level was between 6.0 and 7.4 mg/dl at crossover had
levels well within the range considered "normal," so more stable values of GFR during the one- to two-year followup; one
stringent blood pressure control than traditionally accepted patient who was noncompliant had to go on dialysis.
might exert a favorable effect on renal function. Although 57 In summary, despite many provocative observations, it is not
patients were initially enrolled in the study, only 5 subjects in established whether dietary protein and/or phosphorus restric-
the protein-restricted group and 9 in the control group satisfied tion can slow the progression of chronic renal failure or whether
the requirements of the study. The authors concluded that rates a ketoacid-based regimen confers a therapeutic advantage. One
of progression before and after randomization did not differ, nor or more of the prospective, randomized multicenter trials now
could an effect of dietary protein restriction on progression be in progress might answer these questions.
discerned. Of note, the average intake of protein in the EAA- Let me conclude with a short comment on the progression of
treated group was higher than prescribed and although signifi- renal disease in diabetic patients. A substantial proportion of
cant, the difference from the intake of the unrestricted subjects patients with insulin-dependent diabetes will develop renal
(0.65 g/kgiday versus 0.86 g/kg/day) was not striking. failure. Why some, but not all, patients are at risk is not known.
In summary, an EAA regimen can be effective in controlling A genetic susceptibility to diabetic nephropathy might be im-
the symptoms of chronic renal failure, but any benefit on portant; diabetic siblings of patients with overt nephropathy are
progression of disease remains uncertain. The major advantage more likely to have nephropathy, and black diabetic patients
over a conventional low-protein diet is the greater variety of have a higher incidence of end-stage renal disease when com-
foods, which might make the regimen more acceptable. It pared with white patients. Hypertension can accelerate diabetic
appears to have little or no advantage in terms of improved nephropathy, and effective antihypertensive therapy reportedly
nitrogen conservation when compared with a conventional slows the decline in renal function [59].
low-protein diet. Most investigators believe that microalbuminuria indicates a
Very-low-protein diets supplemented with ketoacids also high risk for the development of clinical diabetic nephropathy.
have been assessed. Barsotti and colleagues examined progres- This observation is of interest because short-term dietary
sion in 31 patients treated with a diet containing 0.5 g/kg of protein restriction reduces protein losses in diabetic patients
high-quality protein and 600 mg of phosphorus/day; all showed with microalbuminuria [601.
a linear decline in creatinine clearance despite dietary protein Data on the efficacy of dietary protein restriction in slowing
restriction [56]. Twelve subjects were then treated for 10 to 15 progression of renal failure are limited. In a recent study of 19
months with approximately 0.2 g protein/kg/day, 300 mg/day of insulin-dependent diabetic patients with persistent proteinuria,
phosphorus, plus a ketoacid-amino acid mixture. Eleven of the conversion from an unrestricted diet (average, 1.13 g protein/
12 had a marked decrease in the loss of creatinine clearance; kg/day) to a diet averaging 0.67 g proteinlkg/day was associated
only one patient continued to lose renal function at the same with a significant reduction in the rate of decline in GFR (from
rate. The same group reported their experience with this 0.61 to 0.14 mI/mm/month) [61]. Slowing of progression was
regimen in a larger number of patients whose renal insufficiency significant even after adjustments were made for differences in
progressed despite therapy with a conventional low-protein blood pressure, energy intake, and glycosylated hemoglobin
 Nephrology Forum: Protein restriction in CRF 337

level. Albumin excretion and its fractional clearance also fell patient discussed had inherited nephropathy, yet apparently she
with the low-protein diet. Another recent study found a similar responded. Barsotti suggested that the ketoacid regimen slows
benefit of a low-protein diet on the progression of diabetic the decline in creatinine clearance experienced by patients with
nephropathy [62]. polycystic disease [66].
In summary, dietary protein restriction has been used by DR. MAmAs: Regarding your data in rats showing that
many investigators to treat uremic symptoms. Our understand- metabolic acidosis accelerates protein catabolism, have studies
ing of the metabolic changes required to maintain nutrition and in humans with chronic renal failure related protein degradation
to monitor compliance has increased rapidly. However, much to the prevailing degree of acidemia? Have any formal studies
work is needed to define how uremia affects protein turnover examined this issue in adults with renal tubular acidosis?
and how changes in lean body mass can be monitored most DR. MITCH: I know of 3 studies showing that correcting the
effectively. Foods designed to be low in protein and phosphorus plasma bicarbonate improved nitrogen balance in patients with
are needed to improve compliance. Such knowledge will be chronic renal failure [67—69]. Children with inherited renal
especially useful if results from prospective clinical trials tubular acidosis have improved growth when they are given
strengthen the results of the reports reviewed. sodium bicarbonate [70]. Finally, Bergstrom and associates
recently showed that the blood bicarbonate level in dialysis
Questions and answers patients correlates directly with the valine concentration in
DR. NIcoLAos E. MADIAS (Chief, Division of Nephrology, muscle; hence, acidosis likely stimulates valine catabolism and
New England Medical Center, Boston, Massachusetts): You limits protein utilization [23].
stated that a decrease in protein intake in the company of DR. MADIAS: Has the effect of respiratory acidosis on protein
ketoacid supplementation can improve renal osteodystrophy on metabolism been examined?
the basis of decreasing the level of PTH and increasing the DR. MITCH: Not to my knowledge.
concentration of 1,25 dihydroxycholecalciferol. Is there any DR. PAUL KURTIN (Director, Dialysis Unit, New England
evidence derived by direct or indirect methods that the severity Medical Center): In contrast to your studies and those of others
of renal osteodystrophy is indeed lessened by utilization of such you quoted [33, 34], growing infants can incorporate exogenous
a dietary regimen? urea into non-essential amino acids. In addition, we have
DR. MITCH: Data from Germany suggest that a low-protein, observed that adults in acute renal failure given parenteral
low-phosphorus diet plus calcium ketoacids is associated with a essential amino acids and adequate calories can have a fall in
reduction in the level of parathyroid hormone in the blood [631. BUN [71]. This finding suggests the incorporation of endoge-
Maschio and colleagues showed that patients fed 0.6 g protein! nous urea into non-essential amino acids and protein anabolism.
kg/day had stabilization of the histology of bone on bone biopsy Would you please comment on this? Second, would you com-
[64]. I don't know of a study of bone biopsy in patients given ment on the studies by Bonomini, which suggest that the longer
ketoacids. a patient is on a low-protein diet, the worse the results of
DR. JOHN T. HARRINGTON (Chief of Medicine, Newton- anemia and nerve conduction studies after dialysis is started
Wellesley Hospital, Newton, Massachusetts): You mentioned [72].
that a decreased dietary protein intake decreases amino acid DR. MITCH: Not only in infants but also in malnourished
oxidation. What is the link between those two steps? adults can urea improve nitrogen balance [73]. On the other
Da. MITCH: I don't think anybody knows. One possibility is hand, in adult patients with chronic renal failure fed an ade-
that the low-protein diet reduces glucocorticoids because glu- quate amount of calories, we find no evidence that urea
cocorticoids influence the activity of the enzyme that breaks nitrogren is utilized to synthesize amino acids. In patients with
down the essential, branched-chain amino acids [65]. acute renal failure, the lower nitrogen intake occurring with
DR. ANDREW S. LEVEY (Division of Nephrology, New En- feeding essential amino acids could reduce urea production and
gland Medical Center): I'd like to return to data from the study hence the BUN. Regarding the initiation of dialysis, the deci-
by IhIe et a! [50]. These investigators demonstrated a lower rate sion generally is based on clinical criteria, and if patients do not
of loss of renal function in patients with moderately severe renal have uremic symptoms, dialysis is not necessary.
insufficiency (mean GFR, 25 mI/mm/I .73 m2) treated with a DR. KURTIN: I interpreted Bonomini's studies as showing
low-protein diet. It was interesting to note that protein intake that if patients are started on dialysis based on a pre-determined
estimated from urinary urea excretion was 0.7—0.8 g/kg/day in BUN, and if they are on a low-protein diet, it will take longer
the low-protein-diet group and 0.9—1.0 g/kg/day in the unre- for these patients to reach the target BUN, while other compo-
stricted-protein-diet group. Thus the dietary intervention ap- nents of the uremic syndrome can progress.
peared to be only a relatively small reduction in protein intake, DR. MITCH: 1 suspect we need careful studies to examine
and the level of protein intake that was achieved was a level that whether dialysis benefits these patients.
is customary in many parts of the world. Would you comment DR. MADIAS: To what extent should the recommendations
on the "dose-response" relationship between the level of for protein restriction in chronic renal failure be changed as a
protein intake and the proposed beneficial effect on delaying the function of age, that is, for children, adults, and the elderly?
progression of renal disease? Also, what about patients who are on high-dose steroid treat-
DR. MITCH: Interestingly, the protein intake slowing progres- ment?
sion in diabetic patients also was approximately 0.7 glkg/day. DR. MITCH: You asked two important questions. Nobody has
Perhaps a response could be achieved without marked dietary studied requirements as a function of the age of uremic patients.
restriction, or it might be that the response is graded, or that Patients given high doses of glucocorticoids, and patients with
patients with certain types of disease won't respond. The the nephrotic syndrome also remain a mystery, at least regard-
338 Nephrology Forum: Protein restriction in CRF

ing their protein requirement. Studies of the safety of low- the BUN and weight are changing (weight being an index of
protein diets in severely nephrotic patients and/or patients total body water), then urea excretion cannot be used to
taking glucocorticoids are needed because this is an important estimate protein intake. If the BUN and weight are stable and
clinical problem. urea nitrogen excretion is consistently higher than the amount
DR. MADIAS: Should one modify the dietary recommenda- of dietary protein nitrogen prescribed, then only two possibili-
tions during periods of stress (for example, intercurrent illness) ties exist: (1) the patient is eating too much, or (2) the patient is
as one does for non-uremic subjects? catabolic or has gastrointestinal bleeding. Distinguishing be-
DR. MITCH: That approach appears rational as long as the tween the possibilities requires careful examination and evalu-
extra protein does not cause uremic symptoms. ation by a dietician. Finally, the test should be repeated.
DR. HARRINGTON: I remain a skeptic about the effect of a low DR. PERRONE: Are there any specific parameters of catabo-
protein intake. I would argue that the most important thing you lism that you would follow up?
did was control the patient's blood pressure. Both experimen- DR. MITCH: During long-term therapy, body weight, serum
tally and clinically, in say accelerated hypertension and in albumin, and transferrin seem to be reliable. We believe the
diabetes, I believe that control of blood pressure is paramount. serum bicarbonate should be kept above 22 mM.
In the studies that have been done about the efficacy of a low DR. BRIAN PEREIRA (Fellow, Division of Nephrology, New
protein intake, has the role of control of hypertension been England Medical Center): What would you consider early
taken into account? I can't imagine that any nephrologist would enough, as far as dietary intervention is concerned? My second
leave patients with blood pressures of 222/120 mm Hg merely to question relates to protein restriction in acute renal failure. Has
carry out the rigorous study that I want. anyone studied whether intervention has any effect on the
DR. MITCH: The definitive study hasn't been done because speed or completeness of recovery in acute renal failure?
blood pressure is always treated. However, controlling blood DR. MITCH: These data do not prove protein restriction slows
pressure alone seems to have the most benefit in slowing progression even though the diet will reduce symptoms. If a
progression to renal failure in patients with accelerated hyper- patient has a rising serum creatinine and a high BUN, then a
tension. low-protein diet should be considered. Regarding low-protein
DR. LEVEY: In the study by Ihle et al, patients in the diets in acute renal failure, attempts to show improved survival
low-protein-diet group lost 2—4 kg in weight on average during from acute renal failure with different types of amino acids or
the first few months of followup. In my experience, this is a low-protein regimens have not proved that survival improves
common finding in patients who follow a low-protein diet. Has [76]. I believe that we need a method of limiting catabolism
anyone carefully studied such patients to determine whether stimulated by other illnesses or renal failure itself. We can
weight loss is the result of loss of body fat, salt, or muscle? replace renal function, but until catabolism can be controlled, it
DR. MITCH: In the patients we treated for many months with seems unlikely that survival time will improve.
a very-low-protein diet, we found no significant change in DR. HARRINGTON: How much time does it take for the
weight on average [32, 57]. Some patients did lose weight, but nutritionist to teach a new patient about a 0.6 g/kg intake, and
we did not examine whether there was loss of bone, fluid, or how often do patients have to see the nutritionist? Please try to
muscle and fat. Lucas et al prescribed half the amount of be as quantitative as possible in your answers.
ketoacids and less than half the amount of protein as others use, DR. MITCH: That depends on the intelligence of the patient
so those results should not be extrapolated to other patients and his or her commitment plus the skill of the dietician.
treated with ketoacids [191. Dieticians in the MDRD study spend quite a bit of time
DR. MADIAS: In your view, what is the importance of changes educating these patients, but in our studies the dietician meets
in insulin and glucagon levels for the abnormalities in interme- with the patient during the initial hospitalization in the clinical
diary metabolism in uremia? research center and monthly thereafter. Each visit lasts 30 to 45
DR. MITCH: We found that uremia is associated with resis- minutes. After about 6 to 8 months, many patients are seen
tance to insulin in terms of protein metabolism [74]. The every other month, and we calculate that they are compliant
importance of glucagon and other hormones in impairing pro- based on urea excretion and body weight [1].
tein metabolism in uremia is less certain [20]. DR. KURTIN: In healthy individuals, as total protein intake
DR. HARRINGTON: What is the impact of gender on progres- falls, energy intake must increase to maintain nitrogen balance.
sion of renal disease? I ask because Lombet et al have reported Do you maintain your patients on diets of 35 kcal/kg? What
that renal insufficiency in the ¾ nephrectomized female rat caloric intake do you recommend for the very-low-protein
didn't progress at the same rate as in male rats [75]. diets?
DR. MITCH: Some studies suggest that women respond to a DR. MITCH: It is difficult to maintain a low-protein diet and a
low-protein diet, but the question has not been examined in high-calorie diet without using supplements of carbohydrate
detail [49, 53]. polymers. This supplement is not too sweet and will increase
DR. RONALD PERRONE (Division of Nephrology, New En- calories. However, we often do not achieve the desired 35
gland Medical Center): When you first place an ambulatory kcal/kg.
patient on a low-protein diet and measure the estimated protein DR. GEETHA NARAYAN (Nephrologist, St. Elizabeth's Hos-
intake using urea nitrogen, how do you distinguish the non- pital, Brighton, Massachusetts): In developing countries like
compliant patient from the patient in negative nitrogen balance? India, renal disease requiring dialysis seems to be just as
At what point are you able to rely on the urinary urea nitrogen prevalent in lower socioeconomic groups as in higher despite
as an indicator of compliance? the fact that these people often consume diets very low in
DR. MITCH: You raise critically important issues. As long as protein, even bordering on protein calorie malnutrition at times.
 Nephrology Forum: Protein restriction in CRF 339

Hence, at least in these populations, other factors seem to DR. LEVEY: Two groups of patients are being evaluated in the
override the effects of a low-protein diet. Would you comment MDRD study; one group has GFRs of 25—55 mI/mm/I .73 m2
on this, and are any data available on protein intake in these (Study A), and the other has GFRs of 13—24 ml/min/l.73 m2
patients? (Study B). The patients in Study B have renal function compa-
DR. MITCH: You raise an interesting point. Perhaps patients rable to that in the patients studied by IhIe et al in Australia. As
in underdeveloped countries don't see a physician until they of the end of 1990, we have randomized 200 patients into Study
have advanced uremia, and they can have secondary illnesses. B, which compares a diet containing 0.575 g/kg/day protein with
The patient discussed today had loss of renal function each time a diet containing 0.28 g/kg/day protein supplemented with
she had a bout of diarrhea and fever. Fortunately, she recov- essential keto acids. This study should enable us to determine
ered. We concluded that an infection along with diarrhea or whether there are different effects of different low-protein diets
fever can impair renal function acutely. If other illnesses in patients with severe impairment in renal function. In addi-
complicate renal failure in a patient from a third world country, tion, in each diet group we are comparing two levels of blood
the patient might not receive medical care until renal failure is pressure: a usual blood pressure goal of MAP (mean arterial
quite advanced. pressure) 107 mm Hg (equivalent to 140/90 mm Hg) and a
DR. MADIAS: What might be the implications of a protein- low blood pressure goal of MAP 92 mm Hg (equivalent to
restricted diet and the provision of adequate energy from 127/75 mm Hg). Hopefully this strategy will enable us to
nonprotein sources on the lipid abnormalities of uremia? determine whether the level of arterial blood pressure affects
DR. MITCH: This question hasn't been examined carefully in the progression of renal disease. In Study A, we have random-
predialysis patients. In patients we have treated, hypertriglyc- ized approximately 400 patients to one of two diet groups, a
eridemia often is present, but an elevated total cholesterol is usual-protein diet containing 1.3 g/kg/day versus a low-protein
unusual. diet containing 0.575 glkg/day. Patients in each group also are
DR. LEVEY: Would you speculate on the mechanism of the randomized to either the usual or low blood pressure goal. This
proposed beneficial effect of low-protein diets and essential study will test the effectiveness of a low-protein diet and
ketoacid supplements in retarding the progression of renal lower-than-usual level of blood pressure in patients with mild to
disease? moderate impairment in renal function. After completion of
DR. MITCH: If the diet works in patients with inherited renal enrollment, the planned duration of followup is approximately 2
disease and in patients with diabetes and/or other types of years.
diseases, a common factor probably is present in all patients.
Such a factor has not been identified. In the patients we treated, Reprint requests to Dr. W. Mitch, Renal Division, Emory University
I have tried unsuccessfully to find a correlation between mean School of Medicine, 136 Clifton Road, NE, Atlanta, Georgia 30322,
blood pressure, systolic or diastolic pressure, and rates of USA
progression. Regarding ketoacids, I don't know whether there
is a special effect; they might permit a lower protein intake. Acknowledgment
Barsotti and associates followed patients whose renal disease This work was supported in part by National Institutes of Health
progressed while they ate 0.6 g protein/kg/day, but this progres- grants ROIDK-40907 and ROIDK-37l75.
sion stopped when the patients began the ketoacid regimen [56].
DR. LEVEY: I agree that the published data comparing rates References
of decline in GFR (measured as clearance of radioisotope-
labeled filtration markers), creatinine clearance, and reciprocal I. MARONI BJ, STEINMAN T, MITCH WE: A method for estimating
serum creatinine are scarce, and that until it is determined that nitrogen intake of patients with chronic renal failure. Kidney mt
27:58—65, 1985
these latter measures correlate closely with the rate of decline 2. GIORDANO C: Use of exogenous and endogenous urea for protein
in GFR, it is important that we include measurements of GFR in synthesis in normal and uremic subjects. J Lab Clin Med 62:23 1—
studies of the effect of interventions on the course of progres- 246, 1963
sive renal disease. I do not agree, however, with the suggestion 3. GIOVANNETTI S. MAGGIORE Q: A low-nitrogen diet with proteins of
high biological value for severe chronic uremia. Lance! 1:1000—
that an apparent linear decline in the rate of decline in renal 1003, 1964
function, measured as either GFR, creatinine clearance, or 4. COLES GA: Body composition in chronic renal failure. Q J Med
reciprocal serum creatinine, justifies using the patients as their 41:25—47, 1972
own "controls" in such studies, rather than using a concurrent 5. Lowtw EG, LAIRD NM, PARKER TF, SARGENT JA: The effect of
control group that does not receive the intervention. Interpre- hemodialysis prescription on patient morbidity. N EngI J Med
305:1176—1181, 1972
tation of changes in the rate of decline in individual patients 6. JOHNSON WJ, HAGGE WH, WAGONER RD. DINAPOLU RP, RosE-
requires assessment of 4 things: (1) carry-over effects from the yEAR JW: Effects of urea loading in patients with far-advanced
previous interval; (2) effects of factors other than the interven- renal failure. Mayo Clin Proc 47:21—29, 1972
tion that could influence renal function; (3) regression to the 7. CONTE G, DAL CANTON A, TERRIBILE M, CIANCIARUSO B, Di
MINNO G, PANNAIN M, Russo D, ANDREUCCI VE: Renal handling
mean [47], and (4) spontaneous breakpoints [47]. In principle, it of urea in subjects with persistent azotemia and normal renal
is possible to assess these effects only if a concurrent control function. Kidney mt 32:721—727, 1987
group is studied or if the experimental group is studied on both 8. RICHARDS P, BROWN CL: Urea metabolism in an azotemic woman
diets in a crossover design in random order. with normal renal function. Lancet 2:207—209, 1975
9. DEFERRARI G, GARIBOTTO G, ROBAUDO C, GHIGGERI GM, lix-
DR. HARRINGTON: I have a question for Dr. Levey. The IANELLO R: Brain metabolism of amino acids and ammonia in
MDRD Study has been going on for a few years. What is the patients with chronic renal insufficiency. Kidney mt 20:505—5 10,
latest information from that important study? 1981
340 Nephrology Forum: Protein restriction in CRF
10. MITCH WE, LIETMAN PS, WALSER M: Effects of oral neomycin
and kanamycin in chronic renal failure: I. Urea metabolism. Kidney
mt 11:116—122, 1977
11. C0TrINI EP, GALLINA DL, DOMINGUEZ JM: Urea excretion in
adult humans with varying degrees of kidney malfunction fed milk,
egg or an amino acid mixture: Assessment of nitrogen balance. J
Nutr 103:11—19, 1973
12. MITCH ME, WiLcox CS: Disorders of body fluids, sodium and
potassium in chronic renal failure. Am J Med 72:536—550, 1982
13. SARGENT J, GOTCH F, BORAH M, PIERCY L, SPINOZZI N, SCHOEN-
FELD P, HUMPHREYS M: Urea kinetics: a guide to nutritional
management of renal failure, Am J C/in Nutr 3 1:1696-1702, 1978
14. YOUNG V: Some metabolic and nutritional considerations of di-
etary protein restriction, in Contemp Issues Nephrol: The Progres-
sive Nature of Renal Disease, edited by MITCH ME, New York,
Churchill Livingstone, 1986, pp 263—283
15. KAYSEN GA, DAVIES RW, HUTCHISON FN: Effect of dietary
protein intake and angiotensin converting enzyme inhibition in
Heymann nephritis. Kidney In! S154—S162, 1989
16. KAYSEN GA, GAMBERTOGLIO J, JIMENEZ 1, JONES H, HUTCHISON
FN: Effect of dietary protein intake on albumin homeostasis in
nephrotic patients. Kidney In! 29:572—577, 1986
17. FAO/WHO/UNO: Energy and protein requirements, in Technical
Report Series 724. Geneva, World Health Organization, 1985, pp
1—206
18. BLUMENKRANTZ MJ, KOPPLE JD, GUTMAN RA, CHAN YK, BAR-
BOUR GL, ROBERTS C, SHEN FH, GANDHI VC, TUCKER CT,
CURTIS FK, COBURN JW: Methods for assessing nutritional status
of patients with renal failure, Am J C/in Nutr 33:1567—1585, 1980
19. LUCAS PA, MEADOWS JH, ROBERTS DE, COLES GA: The risks and
benefits of a low protein-essential amino acid-keto acid diet. Kidney
In! 29:995—1003, 1986
20. MITCH WE, WALSER M: Nutritional therapy of the uremic patient,
in The Kidney (4th ed), edited by BRENNER BM, RECTOR FC,
Saunders, 1991, pp 2186-2222
Philadelphia,
21. HARA Y, MAY RC, KELLY RA, MITCH WE: Acidosis, not
azotemia, stimulates branched-chain amino acid catabolism in
uremic rats. Kidney mt 32:808—814, 1987
22. MAY RC, HARA Y, KELLY RA, BROCK KP, BUSE MG, MITCH WE:
Branched-chain amino acid metabolism in rat muscle: Abnormal
regulation in acidosis. Am J Physio/ 252:E7l2—E718, 1987
23. BERGSTROM J, ALVESTRAND A, FURST P: Plasma and muscle free
amino acids in maintenance hemodialysis patients without protein
malnutrition. Kidney mt 38: 108—I 14, 1990
24. ALVESTRAND A, AHLBERG M, FURST P, BERGSTROM J: Clinical
results of long-term treatment with a low protein diet and a new
amino acid preparation in patients with chronic uremia. C/in
Nephro/ 19:67—73, 1983
25. MAY RC, KELLY RA, MITCH WE: Mechanisms for defects in
muscle protein metabolism in rats with chronic uremia: The influ-
ence of metabolic acidosis. J C/in invest 79:1099—I 103, 1087
26. LI JB, WASSNER SJ: Protein synthesis and degradation in skeletal
muscle of chronically uremic rats. Kidney mt 29:1136-1143, 1986
27. MITCH WE: Uremia and the control of protein metabolism. Neph-
ron 49:89—93, 1988
28. KETI-LEHUT IC, WING SS, GOLDBERG AL: Endocrine regulation of
protein breakdown in skeletal muscle. Diabetes/Metabo/ism Revs
4:751—772, 1988
29. GO0DSHIP THJ, MITCH WE, HOERK RA, WAGNER DA, STEINMAN
TI, YOUNG VR: Adaptation to low-protein diets in renal failure: 52. GUARNIERI GF, TolGo G, SITULJN R, CARRARO M, TAMARO G,
Leucine turnover and nitrogen balance. JAm Soc Nephro! 1:66—75,
1990
30. YOUNG VR: 1987 McCollum award lecture. Kinetics of human
amino acid metabolism: nutritional implications and some lessons. 53. ROSMAN JB, LANGER K, BRANDL M, PIERS-BECHT TPM, VAN DER
Am J Clin Nutr 46:709—725, 1987
31. MOTIL KJ, MATTHEWS DE, BIER DM, BURKE JF, MUNRO HN,
YOUNG VR: Whole-body leucine and lysine metabolism: response
to dietary protein intake in young men. Am J Physio/ 240:E712—
E721, 1981
32. MITCH WE, ABRAS E, WALSER M: Long-term effects of a new
ketoacid-amino acid supplement in patients with chronic renal
failure. Kidney In! 22:48—53, 1982
33. VARCOE R, HALLIDAY D, CARSON ER, RICHARDS P. TAVILL AS:
Efficiency of utilization of urea nitrogen for albumin synthesis by
chronically uremic and normal man. C/in Sci Mo/ Med 48:379—390,
I975
34. MITCH WE, WAL5ER M: Effects of oral neomycin and kanomycin
in chronic uremic patients. H. Nitrogen balance. Kidney Intl 1:123—
127, 1977
35. KOPPLE JD, MONTEON FJ, SHAIB JK: Effect of energy intake on
nitrogen metabolism in nondialyzed patients with chronic renal
failure. Kidney In! 29:734—742, 1986
36. ATTMAN P0, EWALD J, ISAKSSON B: Body composition during
long-term treatment of uremia with amino acid supplemented
low-protein diets. Am J C/in Nutr 33:801—807, 1980
37. MITCH WE, WALSER M: Utilization of calcium L-phenyllactate as
a substitute for phenylalanine by uremic subjects. Metabo/ism
26:1041—1044, 1977
38. MITCH WE, CLARK AS: Specificity of the effect of leucine and its
metabolites on protein degradation in skeletal muscle. Biochem J
222:579—586, 1984
39. MITCH WE, WALSER M, SAPIR DG: Nitrogen-sparing induced by
leucinecompared with that induced by its keto-analogue, alpha-
ketoisocaproate, in fasting obese man. J Clin Invest 67:553—562,
1981
40. MONTEON FJ, LAIDLAW SA, SHAIB JK, KOPPLE JD: Energy
expenditure in patients with chronic renal failure. Kidney Int
30:741—747, 1986
41. SHEMESH 0, GOLBETZ H, KRISS JP, MYERS BD: Limitations of
creatinine as a filtration marker in glomerulopathic patients. Kidney
mt 28:830—838, 1985
42. LEVEY AS: Nephrology Forum: Measurement of renal function in
chronic renal disease. Kidney mt 38:167—184, 1990
43. MITCH WE, COLLIER VU, WALSER M: Creatinine metabolism in
chronic renal failure. C/in Sci 58:327—335, 1980
44. WALsER M: Progression of chronic renal failure in man. Kidney In!
37:1195—1210, 1990
45. WALSER M, DREW HH, LAFRANCE ND: Creatinine measurements
often yield false estimates of progression in chronic renal failure.
Kidney Int 34:412—418, 1988
46. MITCH WE, BUFFINGTON GA, LEMANN J, WALSER M: A simple
method of estimating progression of chronic renal failure. Lance!
2:1326—1328, 1976
47. Modification of Diet in Renal Disease Study, prepared by LEVEY
AS, GASSMAN JJ, HALL PM, WALKER WG: Assessing the progres-
sion of renal disease in clinical studies: Effects of duration of
followup and regression to the mean. JAm Soc Nephrol, in press
48. MA5cHI0 G, OLDRIZZI L, TESSITORE N, D'ANGELO A, VALVO L,
LUPO A, LosCHIAvo C, FABRIS A, GAMMARO L, RUGIU C,
PANZETTA G: Effects of dietary protein and phosphorus restriction
on the progression of early renal failure. Kidney In! 22:371—376,
1982
49. ROSMAN JB, MEIJER S, SLUITER Wi, TER WEE PM, PIERS-BECHT
TP, DONKER AiM: Prospective randomised trial of early dietary
protein restriction in chronic renal failure. Lancet 2:1291—1295,
1984
50. IHLE BU, BECKER GJ, WHITWORTH JA, CHARLWOOD RA,
KINCAID-SMITH PS: The effect of protein restriction on the pro-
gression of renal insufficiency. N Engl J Med 321:1773—1777, 1989
51. OLDRIZZI L, RUGIU C, MASCHIO G: The Verona experience on the
effect of diet on progression of renal failure. Kidney mt 36:S103—
S105, 1989
LUCCHESLI A, OLDRIz2I L, RUGIU C, MASCHIO G: Nutritional
state in patients on long-term low-protein diet or with nephrotic
syndrome. Kidney In! 36:S195—S200, 1989
HEM GK, TER WEE PM, DONKER AJM: Protein-restricted diets in
chronic renal failure: A four year follow-up shows limited indica-
tions. Kidney Int 36:S96—Sl02, 1989
54. ALVESTRAND A, AHLBERG M, BERGSTROM J: Retardation of the
progression of renal insufficiency in patients treated with low-
protein diets. Kidney Int 24:S268—S272, 1983
55. BERGSTROM J, ALVESTRAND A, BUCHT H, GUTIERREZ A: Stock-
holm clinical study on progression of chronic renal failure—An
interim report. Kidney Int 36:Sll0—S114, 1989
 Nephrology Forum: Protein restriction in CRF
56. BARSOTTI 0, GulDucci A, CIARDELLA F, GIOVANNETTI S: Effects
on renal function of a low-nitrogen diet supplemented with essential
amino acids and ketoanalogues and of hemodialysis and free
protein supply in patients with chronic renal failure. Nephron
27:113—1 17, 1981
57. MITCH WE, WALSER M, STEINMAN TL, HILL S, ZEGER S, TUNG-
SANGA K: The effect of keto acid-amino acid supplement to a
restricted diet on the progression of chronic renal failure. NEngIJ
Med 311:623—629, 1984
58. WALSERM, LAFRANCE ND, WARD L, VANDUYN MA: Progression
of chronic renal failure in patients given ketoacids following amino
acids. Kidney mt 32:123—128, 1987
59. PARVING H-H, HOMMEL E, SMIDT UM: Protection of kidney
function and decrease in albuminuria by captopril in insulin depen-
dent diabetics with nephropathy. Br Med J 297:1086—1091, 1988
60. COHEN D, DODDS R, VIBERTI G: Effect of protein restriction in
insulin dependent diabetics at risk of nephropathy. Br Med J
294:795—798, 1987
61. WALKER JD, DODDS RA, MURRELLS Ti, BENDING JJ, MATTOCK
MB, KEEN H, VIBERTI GC: Restriction of dietary protein and
progression of renal failure in diabetic nephropathy. Lance: 2:1411—
1414, 1989
62. ZELLER KR, WHITTAKER E, SULLIVAN L, RASKIN P. JACOBSON
HR: Effect of restricting dietary protein on the progression of renal
failure in patients with insulin-dependent diabetes mellitus. N EngI
J Med 324:78—83, 1991
63. SCHAEFER K, VON HERRATH D, ASMUS 0, UMLAF E: The benefi-
cial effect of ketoacids on serum phosphate and parathyroid hor-
mone in patients with chronic uremia. C/in Nephrol 30:93—96, 1988
64. MASCHIO G, TEssiroRE N, D'ANGELO A, BONUCCI E, Luo A,
VALVO E, LOSCHIAVO C, FABRIS A, MORACHIELLO P, PREVIATO
G, FIAscHI E: Early dietary phosphorus restriction and calcium
supplementation in the prevention of renal osteodystrophy. Am J
C/in Nutr 33:1546-1554, 1980
65. BLOCK KP, RICHMOND WB, MEHARD WB, BUSE MG: Glucocor-
341
ticoid-mediated activation of muscle branched-chain a-ketoacid
dehydrogenase in vivo. Am J Physiol 252:E396—E407, 1987
66. BARSOTTI G: The role of dietary treatment on the progression of
chronic renal disease, in First Taormina Course in Nephrology,
edited by CONSOLO F, BELLINGHIERI 0, SAVICA V, Editoriale
Bios, Cosenza, Italy, 1990, pp 299—313
67. LYON DM, DUNLOP DM, STEWART CP: The alkaline treatment of
chronic nephritis. Lancet 2:1009—1013, 1931
68. PAPDOYANNAKIS NJ, STEFANIDIS CJ, MCGEWOEN MG: The effect
of correction of metabolic acidosis on nitrogen and potassium
balance in patients with chronic renal failure. Am J Clin Nutr
40:623—627, 1984
69. JENKINS D, BURTON PR, BENNETr SE, BAKER F, WALLS J: The
metabolic consequences of the correction of acidosis in uremia.
Nephrol Dial Transplant 4:92—95, 1988
70. MCSHERRY E: Nephrology Forum: Renal tubular acidosis in child-
hood. Kidney In: 20:799—809, 1981
71. KURTIN PS, KOUBA J: Profound hypophosphatemia in the course
of acute renal failure. Am J Kidney Dis 10:346—349, 1987
72. B0N0MINI V, VAGELISTA A, STEFONI S: Early dialysis in renal
substitutive programs. Kidney In: 13 (suppl 8):Sl l2—Sl16, 1978
73. TRIPATHY K, KLAHR S, LOTERO H: Utilization of exogenous urea
nitrogen in malnourished adults. Metabolism 19:253—262, 1970
74. MAY RC, CLARK AS, GOHEER A, MITCH WE: Identification of
specific defects in insulin-mediated muscle metabolism in uremia.
Kidney In: 28:490—497, 1985
75. LOMBET JR. ADLER SG, ANDERSON PS, NAST CC, OLSEN D,
GLASSOCK RJ: Sex vulnerability in the subtotal nephrectomy model
of glomerulosclerosis (abstract). Am Soc Nephrol, 1987, p 23lA
76. MITCH WE, WILMORE DW: Nutritional considerations in the
treatment of acute renal failure, in Acute Renal Failure (2nd ed),
edited by BRENNER BM, LAZARUS JM, New York, Churchill
Livingstone, 1988, pp 743—766
77. SHAH By, LEVEY AS: Applications and limitations of reciprocal
serum creatinine versus time plots (abstract). Kidney mt 35:200,
1989